Gender bias under the microscope

x Sponsored content on Research Live and in Impact magazine is editorially independent.
Find out more about advertising and sponsorship.

I want to start with some interesting facts about the height of CEOs in America: the average American male is 5ft 9in; in the US population as a whole, about 14.5% of all men are six foot or over; yet among the CEOs of Fortune 500 companies it turns out that 58% are six foot or taller.

We haven’t even got to the gender bias element yet but I am sure you will agree that it seems unlikely that when the post of CEO becomes vacant in a large corporation a message is put out suggesting only tall men need apply. Indeed, it seems relatively unlikely that those choosing the next CEO would consider height a relevant factor in the final choice at all.

However, there is clearly an unconscious bias or unconscious prejudice at play that leads those choosing CEOs to prefer tall candidates. The recent debate within the BBC about whether institutional gender bias exists may be explicit gender bias or it may be implicit (unconscious) or possibly both.

The role of unconscious prejudice or unconscious bias in decision-making is increasingly seen as dictating our final choices in everything from choosing one brand over another to finding a partner. Recent academic work shows that our unconscious bias has a more profound and longer lasting impact on decision-making than our conscious bias (though they can, of course, reinforce each other too).

So when Dads4Daughters – a new organisation aimed at raising awareness of gender bias in the workplace – came to Blinc for help, we undertook what is probably the biggest test on unconscious gender bias ever carried out in the UK.

Our test was taken by nearly 8,000 respondents from all walks of life, from factory workers to chief executives, men and women, young and old.

We used our Implicit Association Test or IAT – of which the Harvard IAT and Affective Priming test are the two most famous – developed with our partners Split Second Research to assess unconscious gender bias in a series of different contexts. This is the same test used in academia in the area of racial prejudice to show that even the most liberal minded of us are, in fact, unconsciously a little bit more racist than we would like to think. People tend to show a same-race preference, so white test-takers show a preference for white people, and black test-takers show a preference for black people.

Gender bias is defined for the purposes of this study as: adherence to traditional views of gender roles, especially in terms of working life; a non-conscious belief that certain professions, career routes, personal traits, and roles in the workplace are associated with men and certain others are associated with women.

In the gender bias test, we use around 25 attributes – or primes – words that might traditionally or  stereotypically be seen as male or female, particularly in the workplace. The ‘targets’ in the test are male and female and respondents will need to categorise words as being either clearly male or female – such as girl, boy, man, woman, his, hers – after being exposed to the primes in random sequence.

The degree to which a prime is unconsciously perceived as incongruent with the following male/female target task is the degree to which the task will be slowed down (compared to timing benchmarks taken pre-test). This is all measured in milliseconds.

So what did we learn?

Perhaps the least surprising results were that the more skilled professions, technical subjects, managerial positions and leadership skills are still particularly associated with men at a non-conscious level. 

There is also noticeable dissonance between what a lot of people ‘think’ they feel/think and what they actually think deep down (at the non-conscious level). Many appear to believe they hold more liberal views than they really do. 

Interestingly, associations with females are stronger overall than associations with males. In other words, stereotypes may apply more to women than to men, which raises the question of whether in society we place more constraints on what we expect from women than what we expect from men.

We divided the contexts in which we assessed the possible gender bias into four key areas: professions; career fields; roles in the wider community; and personal qualities. It was clear that professions (such as surgeon, teaching assistant, and so on) and career fields (for example, subjects chosen at college) are far more prone to gender stereotyping than roles (such as leader, follower) or personal qualities (for example, intelligent, gossip).

We also looked at the difference between men and women test-takers in their degrees and areas of gender bias. While there is a high consistency between men and women in their biases (correlation= 0.93 ), there are also clear differences.

Men will tend to subscribe to the female teaching assistant stereotype more than women. Women are more likely than men to stereotype airline pilots as male – as well as politicians – but men are more likely than women to stereotype the manager as male.

Subconsciously, women tend to believe they are more decisive than men, arguably breaking with the stereotype men have of women on this matter. An intriguing feature of the data is that women show unconscious bias as strongly as men do, and sometimes even more strongly than men do (see diagram, especially surgeon, art, fashion, and gossip).

We can also see that, in general, gender bias tends to increase with age or, to put a more positive spin on it, the youngest groups tend to show the least gender bias.

And what about the effect of having children – both sons and daughters? First, we should note that the gender of one’s children does not strongly predict subconscious attitudes.

The only finding here was that those who have only sons show the strongest gender bias in professions but they also tend to think that women have more positive personal qualities. Parents who only have boys adhere to the unconscious gender bias more so than if parents have at least one girl. So, the influence of having a girl is to reduce the impact of unconscious gender bias. Also, only having boys may make parents feel – unconsciously – that girls have more appealing personal qualities!

If we look at different job sectors, then we can see that construction, transport, and finance have the strongest gender bias overall, and this may not be surprising as these fields are visibly male dominated. If we look at what we have termed ‘roles’ (for example, leader, follower, and so on) we see that law has the strongest bias in this area. 

There are sectors of work where there is noticeably low gender bias, such as hospitality, public administration and manufacturing, probably because there is a higher ratio of women working in these sectors.

So, the unconscious stereotyping of what is a male or female role in society is still very strong. According to our research, this prejudice or bias is more noticeable in the world of work, and is much less prevalent in wider society roles and almost absent in personal qualities.

Perhaps most surprising of all is the fact that women are just as guilty as men in adhering – albeit unconsciously – to gender stereotypes. 

Where does that leave us and what should we do?

Younger adults show much less gender bias, so the future looks likely to become even less prejudiced. However, it may be that as experience at work increases, the more one absorbs dominant themes in the workplace, such as gender bias and, if so, we should actively seek to eliminate or reduce such effects.

There may even be an argument for positive discrimination hidden beneath these findings – because if women are actually just as gender-biased as men, they may need artificially to create or demand equality to effect further change. They may need to ‘get over themselves’ to arrive in a truly gender-equal utopia, though to be less gender-biased about it, men need to get over themselves too on this issue.